Paul's fifth anniversary

Both Xconomy and the U-T covered this week’s shareholder meeting. (I’m sorry I couldn’t make it, but it’s been harder to cover the meetings since I moved to Silicon Valley.)

Former U-T writer Bruce Bigelow focused on the five year anniversary of Paul Jacobs being promoted to CEO, while the U-T focused on the share price.

Alas for shareholders, the latter story is far less interesting. Now at $54, the stock has never tested its record $88 close at the end of 1999, at the end of the telecom bubble. While the stock has recovered wonderfully from its post-NASDAQ low of $13, for the past five years it has gone back and forth in a broad trading range of $30-60.

In that regard, it almost exactly mimics my Apple stock in the 1980s, when the correct strategy was to wait for the shares to double to $50, then dump them, and buy them back as they got closer to $25. (This strategy would have only worked twice in the last seven years with AAPL, with the dips of February 2008 and December 2008 legitimate buying opportunities.)

While 2010 provided record net income, sales have essentially been flat for the past three years. The stock price is highly dependent on the growth multiple, and the U-T quoted an analyst who’d downgraded the stock due to slowing growth.
So while the industry is continuing to expand its use of mobile data services — the driver of profitability for QCOM in this century — it’s not clear how much of that will accrue to the share price and its shareholders. The pressure on QCOM over the past few years has been the commoditization of even high-end handsets, pushing down wholesale prices and thus the basis for QCOM’s royalty payments.

As Bigelow recounts, Jacobs fis has a sound strategy for expanding Qualcomm’s influence and reach in the wireless industry, beyond cellphones and (with the Atheros purchase) beyond cellular to Wi-Fi and other forms of wireless communications. Still, it’s not clear whether these efforts will grow the company or merely replace shrinking revenues in existing business.

In this regard, Qualcomm is looking like any other mature tech company, like IBM, HP or Intel. A retailer or other low tech company can usually count on the same revenues year in, year out — but Moore’s Law tech companies must accepted that the old products will decline in price and thus must be replaced by upgraded or entirely new products.

Perhaps the world’s largest semiconductor company offers a ray of hope: Intel revenues were declining from 2007-2009 until things rebounded in 2010 with revenues up 24% and record earnings. According to its January earnings call, Intel benefitted from exposure to rapidly growing product categories and geographies. It remains to be seen whether Qualcomm’s diversification will bring it similar dividends in 2011.

San Diego to WSJ: what about us???

The front page of the Bay Area edition of the Wall Street Journal this morning proclaimed: “Wireless Industry Calls Valley Home.” The inside headline said: “Phone Makers’ New Area Code: 650.” For members of the San Diego telecom industry, those should be fighting words.

The premise of the story:

According to IDC, sales of smartphones are expected to grow 39% world-wide this year from 2010 to 421 million units. More than 40% of those devices will run on operating systems developed within 10 miles of each other in Silicon Valley.

The article by veteran tech reporter Yukair Iwatani Kane presented a oversimplified and distorted version of the US wireless industry. It played up Sony Ericsson — the has-been cellphone marker in 2010 to 6th in global market share and off the map in smartphones. It never mentioned chips at all — nor Qualcomm, the largest cellphone chip maker or its San Diego hometown.

However, Kane can’t be held responsible for the most glaring error, that of the headline: the iPhone, its OS and app store are developed and run in the 408 area code.

Still, if the measure of mobile phone leadership is software platform market share (arguable but plausible), Silicon Valley is out in front. It didn’t have to be so: Seattle (through Microsoft) had many chances but blew it, and London was ahead for a decade (due to Nokia’s investment in Symbian Ltd.) until Nokia started to fold.

This of course is about the secular shift in cellphones: it’s not about the radios and networks, it’s about the software, platform, application and the Internet. (Intel’s dreams notwithstanding, the chips are all ARM licensees which means Qualcomm must fight relentlessly to gain and maintain market share.)

If the fight is over software, then San Diego will play a decreasing role in the growth of the wireless industry. When I helped start the SDSIC in 1993, we were concerned about the region’s ability to support local software companies, and modeled some of our practices after Silicon Valley forebears. Despite our hopes, the region’s software industry never grew all that much — certainly trailing Silicon Valley, Seattle and several other cities.

UCSD and Qualcomm alumni are starting firms, but the software engineering and VC talent will remain concentrated elsewhere in the state.

Qualcomm itself would rather switch than fight. Under Jacos fils, it’s become less interested in San Diego and is expanding in the Bay Area to get local design wins and tap its software and Internet expertise.

Effect of zombie portfolio on LTE royalties?

The WSJ reports that the liquidation of Nortel is moving on to its portfolio of 4,000 telecommunications patents, worth as much as $1 billion.

The most strategically valuable are those related to LTE, given that the Canadian firm was aggressively developing 4G technology before it went bankrupt two years ago.

The WSJ listed four telecom firms as likely bidders — Apple, Google, Huawei and ZTE — all firms relatively light on 3G and 4G patents.

However, the article mentions as a possible bidder only one of the four major 3G patent holders: Nokia, Ericsson, Qualcomm and InterDigital. The latter is mentioned in the same breath as Intellectual Ventures, Nathan Myhrvold’s well known Silicon Valley patent troll:

Closely-held Intellectual Ventures and InterDigital use patents for offensive purposes, licensing them as broadly as possible and asserting them in infringement suits against companies that refuse to take a license.

A possible rival bidder is RPX, a “defensive patent aggregator” that lists Google, HTC, Huawei, Nokia, RIM and Samsung as members. For obvious reasons, Qualcomm is not a client of the company.

Without knowing where the patents will end up, it’s impossible to predict their impact on Qualcomm’s QTL division and its IP-based business model.

However, with two rare exceptions — Broadcom and Nokia — large patent portfolios in the hands of other telecom companies have had no significant impact on the QTL business. Qualcomm has managed to cross-license patents with its customers (including more than 15 years with Nortel) without jeopardizing its royalty rate.

If the patents go to Apple, Google or one of the Asian makers, I don’t think it will impact Qualcomm‘s royalty rate. (Instead, Qualcomm’s pricing power will depend on the relative strength of its LTE portfolio vs. its 3G or cmdaOne holdings.)

I think the story is different if the patents are acquired by IV or InterDigital. Either might choose to sue Qualcomm’s LTE chips for infringement, and — unlike Nortel, Samsung or even Broadcom — they lack their own hostages that QTL can threaten with its patent portfolio. Still, I think Intellectual Ventures is a far more serious threat to QCOM than InterDigital.

In 1993, InterDigital sued Qualcomm (and was countersued) over 2G CDMA patents. The upshot was that Qualcomm paid InterDigital a flat $5.5 million settlement while InterDigital customers paid royalties for use of Qualcomm’s patents. InterDigital’s has been settling with smaller firms, but lost a major case against Nokia in 2009.

InterDigital stock has doubled in less than five months, but is still trading an discount to Qualcomm’s. They have an incentive to rebuild their patent portfolio to strengthen their hand in 4G licensing.

Given they cut their R&D back by 30% in 2009 — and their 2010 quarterlies suggest that R&D remains cut — I don’t see how InterDigital could afford to buy even one of the six portfolios. It’s also not like them to partner — the don’t need a license to patents but the right to assert — but perhaps they could presell rights to the patents they buy to existing customers to help pay for the cost.